metadata
dataset_info:
features:
- name: image
dtype: image
- name: text
dtype: string
- name: format
dtype: string
splits:
- name: train
num_bytes: 3827972504.0879993
num_examples: 88276
- name: validation
num_bytes: 454435527.32
num_examples: 10385
- name: test
num_bytes: 228361839.472
num_examples: 5192
download_size: 4496694222
dataset_size: 4510769870.879999
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
- split: test
path: data/test-*